Fixing the Power Grid

Oct. 17, 2007 — -- Large-scale power storage is crucial to our energy future: the Electric Power Research Institute, the U.S. utility industry's leading R&D consortium, says that storage would enable the widespread use of renewable power and make the grid more reliable and efficient. Recent announcements by utility giant American Electric Power (AEP), based in Columbus, OH, suggest that grid storage technologies are finally ready for commercial deployment in the United States. Last month, AEP ordered three multi-megawatt battery systems and set goals of having 25 megawatts of storage in place by 2010, and 40 times that by 2020.

"That was a dream four or five years ago; now it is happening," says AEP energy-storage expert Ali Nourai.

The AEP system uses a sodium-sulfur battery about the size of a double-decker bus (see below), plus power electronics to manage the flow of AC power in and out of the DC battery. Though new to the United States, the system has been used at the megawatt scale in Japan since the early 1990s; the battery was produced by NGK Insulators of Nagoya, Japan.

Nourai says that AEP and other U.S. utilities gained confidence in the economics and reliability of storage thanks to a demonstration project in Charleston, WV, where AEP installed a large battery system in June 2006. In Charleston, peak demand in both summer and winter had overloaded transformers at local substations, causing blackouts. Rebuilding the substations to accommodate more power could have taken as much as three years. Instead, AEP spent just nine months installing a battery system that charges when demand for electricity is low and can deliver up to 1.2 megawatts for seven hours when demand peaks.

Two of AEP's new projects are slightly larger two-megawatt, seven-hour battery systems designed to provide similar quick fixes in areas with power-reliability problems. A battery in Milton, WV, for example, will provide backup electricity for customers in areas prone to blackouts from a weak power line. "When there is a blackout, the battery will pick up as many people as it can and continue to feed them," says Nourai. "They will not even know there was a blackout." The battery will postpone Milton's addition of a new substation and a high-voltage transmission line by five to six years.

When AEP decides to make more permanent upgrades to substations or completes construction of a new power line--a process that can take five or six years--it will simply move the nearest backup battery to another choke point. "It can be lifted with a forklift and loaded onto a flatbed truck," says Nourai. "Within a week we can have it up and operational at another site in our system."

Richard Baxter, author of Energy Storage: A Nontechnical Guide and chair of a conference held last week in New York City on investing in storage, says that AEP's new projects are a "good litmus test" for the industry. "Storage technologies are emerging as a viable, commercial-level product," Baxter says.

The emergence of a grid storage market is drawing in new battery developers. These include Firefly Energy of Peoria, IL, which is using high-surface-area nanostructured electrodes to revive lead-acid technology, and lithium battery developer Altair Nanotechnologies, based in Reno, NV. In June, multinational utility AES agreed to buy an unspecified number of Altair's batteries; CEO Alan Gotcher says that Altair will deliver a one-megawatt, 15-minute prototype by the end of this year.

AEP, meanwhile, is exploring a potentially more transformative role for storage: turning the ever-shifting power output of renewable resources such as wind and solar power into steady, dependable energy. The company plans to connect its third two-megawatt battery system to a group of wind turbines at an as-yet undetermined site. Nourai says that the goal is to learn whether batteries can smooth out short-term fluctuations in power flow from the turbines. If they can, utilities should be able to absorb larger levels of wind power on their grids.

But Nourai says that AEP also wants to determine whether storing wind energy can boost its value. There are at least two ways that this could happen. Wind energy produced at night could be stored for delivery during peak hours of the day, when the price of electricity spikes. And if the power delivered by wind farms were more predictable, it would be more profitable. When an independent generator such as a wind-farm operator sells to power distributors, it must promise to deliver a certain amount of power at a certain hour. While the details vary greatly in different regional and national power markets, wind-farm operators can be penalized if they fail to meet their commitments because the wind didn't blow as hard as expected. Systems that store a fraction of a wind farm's output when the wind is blowing can eliminate most of this risk.

Nourai notes that Japanese utilities are already installing energy-storage technologies to make wind power more reliable and profitable, thanks to government incentives that cover one-third of the cost of the storage system, and to the wider spread between Japan's day and night electricity prices. Nourai believes that NGK, which can currently produce 90 megawatts' worth of sodium-sulfur battery systems per year, is considering constructing a second factory to meet the resulting demand. Meanwhile, a study completed this year by Sustainable Energy Ireland, Ireland's energy-policy agency, concluded that time-shifting storage projects might already be profitable in Europe.

However, an expert panel assembled by the Electric Power Research Institute last year judged that storage costs needed to drop below $150 per kilowatt-hour to make such time shifting economically attractive in the United States; a report issued by the institute this spring estimates that systems employing NGK's sodium-sulfur batteries cost $300 to $500 per kilowatt-hour. That cost differential has fueled recent interest in solar-thermal-power plants that capture renewable energy in the form of heat, which is easier to store than electricity.